Subjects were 70 Wistar rats showing either low preference for aversive alcohol solutions or a high preference induced by hypothalamic stimulation. Experiments 1 and 2 showed that a large lithium chloride injection.(3 meq/kg) suppressed alcohol intake only if alcohol was tasted. Pairing lithium contiguously with water or intubed alcohol failed to reduce subsequent alcohol intake despite the concurrent presence of high serum lithium levels. In Experiments 3 and 4 a series of seven lithium injections increased rather than decreased alcohol intake if lithium was allowed to accumulate in the blood and brain during alcohol exposure while the transitory sickness associated with each injection was prevented from association with the taste of alcohol. When sickness was allowed to occur during alcohol exposure a suppression of intake resulted after two injections. Contrary to current interpretations these results suggest that the suppression of voluntary alcohol intake by acute and chronic lithium administration is due to a learned taste aversion rather than to a pharmacological mechanism specific to alcohol;